Commentary: Are the NWRL tunnels too narrow and too steep?

Posted: March 24, 2013 in Transport
Tags: , , , ,

The recent article by Sandy Thomas made numerous criticisms of the current government’s policy choice on heavy rail in Sydney, but ended by making a plea with the government to build the North West Rail Link’s tunnels at a width and gradient sufficient to allow Cityrail’s existing double deck trains to pass through it. Not doing so, Mr Thomas argues, would be akin to repeating the mistake of 1855 where different rail gauges were set for different states, forever preventing any movement of trains from one network to another. While his other arguments went so far as being wishful thinking or even approaching conspiracy theories, this is the one point which has garnered the most public support and is the most likely element that the government would change, if it changes anything at all.

The Northwest Rail Link will include a new railway from Epping to Rouse Hill, plus a retrofitted Epping to Chatswood Line. Click on image for higher resolution. (Source: NWRL EIS - Introduction, page 1-3.)

The Northwest Rail Link will include a new railway from Epping to Rouse Hill, plus a retrofitted Epping to Chatswood Line. Click on image for higher resolution. (Source: NWRL EIS – Introduction, page 1-3.)

I am still undecided as to what I think about this move. It does concern me that this will create 2 separate systems, one of which will eventually go as far as Hurstville, Cabramatta, Lidcombe, and Rouse Hill, running exclusively single deck trains. Will these corridors, or even part of these corridors still be using single deck rolling stock 50 or 100 years from now? The simple answer is we just don’t know, so the safe thing to do it to build them with the ability to convert them in future, just as part of the current network is being converted now.

This plan appears even more absurd when you realise that the savings from the NWRL tunnels will be offset by the cost of converting the line between Epping and Chatswood for single deck trains. In other words, it will cost more than creating a new line with just double deck trains.

Countering these are 2 points in favour that are worth considering. First is that if a Second Harbour Crossing is built under the Harbour, then narrower and steeper tunnels will result in much more significant cost savings than they would on just the NWRL portion. Second, there is nothing to say that double deck trains in the future could be small enough and powerful enough to fit into the narrower tunnels and make it through the steeper gradients. In fact, a cynic could argue that since the tunnels are only barely small enough to prevent existing rolling stock from fitting though them, that this was purposely done to prevent only current rolling stock from using it, rather than any potential future rolling stock. It also deserves remembering that the Epping to Chatswood Line’s tunnels were built too steep for some Cityrail trains. So this is not necessarily as much of a long term line in the sand that the different rail gauges of 1855 were (something which can never really be overcome).

At the heart of all of this is whether an independent single deck metro is the best way to run the NWRL. I think it is, on both cases. An independent line will allow the government to escape the bottomless money pit that Railcorp has turned into, perhaps even result in driverless trains that can achieve very high all day frequencies while allowing the cost savings to be directed at better staffing at stations or more roving security guards/police rather than drivers and train guards (responsible for telling you to stand clear of the doors rather than any tangible security) which are redundant given today’s modern technology. A single deck metro is also appropriate for the global economic corridor that the NWRL will cover and its high turnover passenger nature, preventing dwell times from ballooning out and thus maintaining a high level of on time running.

Making the tunnels narrow and steep guarantees that the NWRL will end up as an independent metro line. It prevents any future government from having second thoughts on the matter. So clearly, if you disagree with how the NWRL should operate, as Mr Thomas does, then the tunnel’s decision is the biggest point of focus for you. Therefore, to justify the tunnels as planned, you need to have no doubt in your mind that this is the best option to take.

And this is why I’m still undecided. I have little doubt that an independent, single deck metro line is the best way forward on the NWRL, but that is not a complete absence of doubt. So while I remain open to the decision to build narrower and steeper tunnels, I am hesitant to endorse it. But I do find myself currently leaning towards it.

  1. Simon says:

    Privatise Railcorp I say, then the benefits are systemwide.

  2. moonetau says:

    Which suburban rail networks are privatised? Which are state owned?
    Which have the highest farebox return? Which have the lowest maintenance costs? And so on…

    I support the Ecotransit position: second harbour rail crossing via the eastern lanes of the Bridge.
    Much cheaper than a deep harbour tunnel, in term of upfront cost.

    A year ago a NSW Parliamentary committee report into the cost of rail infrastructure made a number of recommendations.

    One of them was to establish a “Cost Estimating Centre” See Part E p.12.

    Mr O’Farrell also promised a greater level of transparency with regard to government expenditure.
    I think it would be a good idea if TfNSW could give the public all the information that it has (notwithstanding matters of commercial in confidence) available to it.

    For example do single deck metros really have a greater line capacity under modern signalling systems than triple door double deckers? What are the levels of waste in Railcorp (too many staff with too little to do?)? Have the rorts and corruption brought to public attention in the last few years been successfully dealt with? Can they be dealt with without privatisation? If single deck trains can be driverless, why not double deckers? What will be the real cost of a12km deep harbour tunnel and associated stations? What about an 8km shallow tunnel and Bridge option?

    My gut feeling is that the current government is making decisions based not on cost and convenience to the commuter alone but on another, probably political agenda.

    Metros are for densely populated cities not for places like Sydney. Will people from the North West really tolerate standing for even half of the journey to the CBD,
    AND having to change at Chatswood, again without the guarantee of a seat?
    We know that having to stand for 40 to 50 minutes is not a viable option in the 21st Century.

    Larger tunnels will add little to the capital cost so in my opinion that should be the option that is pursued.

  3. Lots of good questions there, many of which I’d like to know the answers to as I don’t know myself.

    As to the driverless on DD trains, there shouldn’t be anything stopping that. I think the NWRL should be single deck and driverless, but came to those 2 conclusions independently. If driverless (or even Automatic Train Operation with someone in the drivers chair supervising) works on the NWRL, then I’d like to see it expanded to the Eastern Suburbs and Illawarra Line, which is another independently operating line. The key issue is independent operation, not single vs double deck.

    I also reject that people from the Northwest will have to stand any longer than train passengers in Sydney already stand. If they get on at the first station on the line, then they will get a seat, as will most in the furthest out stations, so no one will spend an hour from Rouse Hill to the CBD on their feet. Many will also not be going all the way to the CBD – 42% will get off before the Harbour Bridge, 33% will get off by Chatswood.

    Opponents of the single deck metro concept for the NWRL take too much of a CBD-centric view of Sydney (then often also criticise it for not going all the way into the CBD). The reality is that this new line will eventually go to a dense employment centre (the CBD), but pass through a dense employment corridor (the Global Economic Corridor) on the way. This means there will be a high turnover of passengers, which also means those standing passengers on the train get priority over recently vacated seats ahead of passengers boarding at that station. Another reason passengers won’t be standing as long as has been claimed.

  4. moonetau says:

    Might this happen to Sydney?
    “Jobs shift from suburbs to CBD”

    I know that a lot of the published research suggests that the Global Economic Arc will be a source of jobs growth but I wonder for how long? (I’m thinking 20-30 years into the future). And if many of them are going to and from Sydney Airport and thence to other cities / CBDs …

    Yes there is evidence to show that there is strong jobs growth outside CBDs. But the trains are still packed in the peak aren’t they? Especially the western line and it it clear that we need more rail capacity. Indeed it is universally acknowledged that we will need a second harbour crossing at some point in the not too distant future. I’m still not convinced that a single deck NWRL / Metro can move more people more cheaply in relative comfort between the North West and Eveleigh / Redfern / Airport *under* the harbour then *above* it.

    Another thing to consider is growth itself: employment and population. What if it stalls or is limited, as it has to one day.

    A further consideration is the price of fossil fuels, alternative fuels and the place of grid connected mass transit. For me it is the latter which we must work on.

  5. The Global Economic Corridor is closer to where the people live and is more affordable for both businesses and residents. Remember – almost half of Greater Sydney’s population is in Western Sydney, and soon that will be more than half.

    The Sydney CBD has, is, and will continue to be important, but is unlikely to increase its 15% jobs share by more than a few percentage points in coming decades. And that is the beauty of the NWRL plan – it connects orbital journeys from non-CBD origin and destinations (e.g. Rouse Hill to Macquarie Park) as well as trips into the CBD. It will provide the CBD capacity improvements required to reduce the congestion you cite, but it also provides a frequent trunk service through North West Sydney that will be supported by a network of feeder buses, creating the grid that you also desire.

  6. JC says:

    I agree with most of your arguments about the best way forward on this. Flexibility for the future has to be the the best solution, even if it is sensible to do something different now.

    I do however take issue with your continuing references to “Global economic corridor” and the need for a metro to serve it. GEC is planner government hype that shouldn’t distract us. It is admirable policy i.e. reverse the current imbalance by shifting housing east and jobs west – but even if the policy succeeds, employment and population density along the “GEC” will not forseeably ever reach the levels required to support a full scale Singapore/Hong King/London/Paris/New York style metro. Without a major new approach to densities it will always be “suburban” and best served by comfortable (i.e. with seats) commuter trains at reasonable intervals supported by feeder and cross-town buses at high (metro-style) frequency.

    You are right that the people of Western Sydney deserve better public transport to rid them of car tyranny – but an uncomfortable metro that will only be economic to run on hourly or half-hourly frequencies won’t deliver. What we need is convenient and frequent (0-10 min) services, and the mode dictated by density.

    For Western Sydney this is likely to be bus with some busway for LRT. For the central core (i.e. the box bounded by the coast, the harbour/Parramatta River, Duck Creek and Georges River/Botany Bay) it will be rail-based because of the density, but even then probably LRT except for the most heavily trafficked corridors (such as City-Parramatta). Newcastle (UK), Manchester and the big German urban areas show that you can deliver metro operations/performance with LRT technology for medium density areas at a tiny fraction of metro costs – and often using existing assets.

  7. Thought says:

    But, the Global Economic Arc between Chatswood and Macquarie Park (and to its west to Castle Hill), is the one of the least densely populated areas of Sydney.

    The State Government’s newly released urban strategy itself shows that other suburbs and regional centres across Sydney will deliver jobs and housing, this includes the poorly connected Rhodes and Sydney Olympic Park (which currently requires changing trains to reach from most areas in Sydney). Both areas are increasing their residential densities, and combined are likely more significant than Norwest.
    The most densely populated areas along the global economic arc are Chatswood/St Leonards and soon Macquarie Park. These residents will most likely be travelling in one direction to get to work and that’s east/south.

    This ‘metro’ from the Hills to the CBD to Hurstville doesn’t make sense, if you wanted to move people from the most densely populated areas, all the time.

    Metro style services on the NWRL during the day won’t make sense either. These areas are not destinations in their own right during the day, with the exception of Macquarie University.
    For shopping trips, I don’t think many will hop on a train to go to Castle Hill to shop when it’ll be easier to drive and park. Presumably they’ve done studies, but who will be utilising the metro service in non peak times to make it viable?

  8. Simon says:

    Melbourne and Auckland are privatised and have gone forward. Perth is public owned and has gone forward. ADL/WLG/SYD/BNE have languished in public ownership.

  9. Simon says:

    33% of NWRL passengers exiting the rail system by Chatswood seems a very high proportion. Seems unlikely. The density of the employment is relatively low – not really what is needed for rail.

    As for the comments about “everyone says we need a new harbour crossing”. Well we only need the new harbour crossing because of the NWRL. No NWRL, no new harbour crossing needed.

  10. There’s been a lot of arguments against the Global Economic Corridor, but none are supported by the facts. The reality is that more Hills residents worked along the Global Economic Corridor 12 years ago than in the CBD[1]. I imagine this proportion could be much higher today.

    And the employment centres along this corridor are dense ones, measured as jobs per Hectare: North Sydney (369), St Leonards (107), Chatswood (271), Macquarie Park (69), Castle Hill (53), and Norwest Business Park (44)[2]. These figures are from 2006, and pre-date a rail connection to Macquarie Park. Adding rail will allow much higher employment densities, more like the first 3. Compare these to the Sydney wide average of 1.6 jobs/Ha and consider that only the CBD, Parramatta, and Bondi Junction have densities above 100.

    Nor is the argument that the NWRL will only have “hourly or half-hourly frequencies” – the Epping to Macquarie Line (which will form part of the NWRL) already runs at 15 minute frequencies. Again, the facts do not support this argument.

    The reality is that the best way to get people out of their cars is to create a frequent all day network based on a trunk rail line to provide high capacity which is then supported by frequent feeder buses. That is exactly what the NWRL will do.

    But if you base your argument around gut feelings, rather than facts, then yes the NWRL is a piece of lunacy that will be a waste of money which could be spent somewhere else to better effect.


  11. I think 33% is plausible based on the facts. More Hills residents work along the Global Corridor than the CBD (7.6% vs 7.3%, in fact it’s 12.3% vs 7.3% if you include Castle Hill – and that excludes Norwest[1]). The argument that they are low density also doesn’t hold up when you look at the facts – Chatswood has an employment density of 271 jobs/Ha, Macquarie Park is 69, Castle Hill 53, and Norwest 44[2], compared to a Sydney wide average of 1.6 jobs/Ha. These figures are from 2006, and Macquarie’s density would have only increased since it got its rail line in 2009, Castle Hill and Norwest will too once they get their’s.

    I’ve based my judgement on the facts, but it concerns me that you seem to have based your’s on gut feeling. Please tell me that’s not the case – what facts have you used to back up your conclusion?


  12. Well put.

    As I understand it, the Perth bus system is owned and planned by the government but operated privately (the same model being proposed for the NWRL). Happy to be corrected on that though.

  13. Simon says:

    There are limited facts available, but I’m troubled by the presently low number of people using the 619 bus to reach the Macquarie destinations. I know there is a low number of people using it because it operates infrequently – 4/hr in peak. I don’t think that suddenly having a train available will see too many people change from car to driving – that would only work if your workplace is near the station or has limited parking.

    Now this could change in time, particularly if employment density in this area increases and car parks per job reduces, but PT gaining market share in that way is not really a function of the rail line in that case but it is a function of the other concerns, namely reasonable town planning.

  14. Simon says:

    I was only referring to rail systems above. All the cities referred to have fully privatised bus systems except Sydney and Brisbane. Hobart still has public buses and I’m not sure about Canberra and Darwin, but I think they have public ownership.

  15. Simon says:

    Further to my last post, I say what is important are things such as walking distances, catchments, journey times, parking availability, fares, frequency, reliability and comfort. The train line will only improve the last three points. Likely, many people will actually be left with a worse service needing longer walking distances, but some people will be better off, particularly those that do not need to interchange at either end of the train trip.

  16. moonetau says:

    Can I go back to Bambul’s original question? Are the NWRL tunnels too narrow and too steep?

    Which tunnels?
    Between C Road and Epping? I thought the ruling gradient was quite easy (unlike the Epping to Chatswood?)

  17. […] Commentary: Are the NWRL tunnels too narrow and too steep? […]

  18. RichardU says:

    Berlin has an encouraging 70% farebox recovery percentage AND it has integrated fares with only 3 zones. Berlin’s ticketing system does not have a complicated, costly, risky Opal type ticketing system. One suspects taking the cost and complexity out of ticketing helps with the Farebox recovery figures.

    Where is dwell time a bigger issue? On trains or buses? Where is the current government “solving” the problem at the cost of having passengers stand in a “modern” metro. It is not a problem on Berlin buses, you enter and leave buses as you please through any door.

  19. Richard, I don’t think cost recovery has anywhere near as much to do with the method of fare calculation chosen as much as it has to do with population density.

    But if you do want to look at zonal vs distance based systems, then according to your link the 7 systems with the highest cost recovery percentages all use distance based fares. And this includes the only 6 systems that operate without government subsidy (i.e. at a profit rather than a loss).

    Not that this is all that relevant, as the reason for this high cost recovery has more to do with these being very dense Asian cities which lend themselves well to mass transit systems.

    All door boarding would be a great addition to Sydney. Hopefully it will be introduced once Opal has been rolled out (given that there are Opal readers on all doors), much as San Francisco recently allowed all door boarding on its buses.

  20. RichardU says:

    “And the employment centres along this corridor are dense ones, measured as jobs per Hectare: North Sydney (369), St Leonards (107), Chatswood (271), Macquarie Park (69), Castle Hill (53), and Norwest Business Park (44)”

    Meanwhile the Urban Activation Precinct plan for the Epping Junction envisages the 51,000m2 of commercial space being REDUCED by 30-47,000m2 despite the proposed new floor space allowance zoning being clustered around the station, unlike Macquarie Park which has much lower densities and hectares of car parks which are all still full.

  21. […] Commentary: Are the NWRL tunnels too narrow and too steep? […]

  22. mich says:

    In what way has Melbourne “gone forward” ? Melbourne’s trains are rubbish. Slow, infrequent, expensive, dirty, unreliable, it’s like the 70’s all over again in Melbourne.

  23. Simon says:

    Melbourne’s trains have gone forward in the sense of 150% patronage growth over about a dozen years. Sydney’s patronage growth has only been about 20% in a similar time frame, and less money is being spent in Melbourne.

  24. […] narrower and steeper tunnels, which force the new line into being run completely independently from the rest of the network, […]

  25. RichardU says:


    I think you will find Melbourne started from a much lower base of patronage. It is easier to get a large percentage use figure in such cases.

  26. […] Commentary: Are the NWRL tunnels too narrow and too steep? […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s